Why volunteers outperform paid workers on some tasks

An unpaid moderation team maintains a reference archive for six years. A small honorarium is introduced to acknowledge the work. Within two quarters, contributions are being logged, compared, and negotiated, and total output has fallen — not because anyone stopped caring, but because the question being answered changed.

What everyone sees

Money is assumed to be additive: whatever reason existed before, payment supplies an extra one on top. In some settings it substitutes instead. Once an amount is attached, the activity is evaluated against that amount, and a small payment is a weaker case for effort than an unpriced commitment was.

What is actually happening

Deci, Koestner and Ryan’s meta-analysis of experiments on extrinsic reward found that tangible rewards contingent on task engagement reduced measures of intrinsic motivation for activities that were interesting to begin with — a conclusion Cameron and Pierce disputed on coding and moderator grounds, in an exchange that remains unresolved. Gneezy and Rustichini’s field and laboratory results give the sharper form: small payments produced worse performance than no payment at all, with performance rising again only once the amount was large. Ariely, Bracha and Meier add the signalling channel: paid effort is no longer unambiguous evidence of commitment, so the reputational return on doing it well is lower for the worker as well as less legible to observers.

Why it stays hidden

The substitution hides because the intention is generous and the amount is modest. A token payment is offered as recognition, not as a wage, so nobody expects it to function as one. It functions as one anyway, since the recipient now has a figure against which to judge whether the next hour is worth giving.